Microbiology (1 9961, 142,2847-2856
Printed in Great Britain
Transcriptional activity of the symbiotic
plasmid of Rhizobiurn etli is affected by
different environmental conditions
Lourdes Girard,l Brenda Valderrama,’ Rafael Palacios,’ David Romerol
and Guillermo Ddvilal
Author for correspondence: Lourdes Girard. Tel:
e-mail : [email protected]
Departamentos de
Genktica Molecular’ and
Ecologla Molecula r2,
Centro de lnvestigacidn
sobre Fijacidn de
Nitrbgeno, Universidad
Nacional Autdnoma de
Mkxico, Ap. Postal 565-A,
Cuernavaca, Morelos,
Mexico
+ 52 73 175867. Fax:
+52 73 175581.
Global patterns of transcriptional activity of the symbiotic plasmid (pSym) of
Rhizobium etli were studied under a variety of environmental conditions,
including some relevant to the symbiotic process. 32P-labelledsingle-stranded
complementary DNAs synthesized from total RNA were used as hybridization
probes against an ordered collection of cosmid clones that covered the whole
pSym. Our results showed that, under aerobic conditions, discrete regions of
the pSym are differentially transcribed depending on the carbon and nitrogen
sources employed. In general, poor carbon or nitrogen sources allowed greater
expression than rich ones. Time-course experiments with the nod gene inducer
genistein led us to the identification of new regions responsive to this
f lavonoid. Widespread transcription was observed during microaerobiosis, but
not under aerobic conditions, indicating that oxygen concentration is a major
effector of transcriptional activity in the pSym. This response was reduced, but
not suppressed, in a nifA mutant, indicating the location of regions whose
transcription may depend on other oxygen-sensitive regulators. During
symbiosis, almost the entire pSym was actively transcribed and the
transcription pattern was similar to that observed during microaerobiosis. The
experimental approach described allowed the identification and localization of
specific regions in the pSym whose expression depends on defined
environmental stimuli.
Keywords : Rhixobium etli, symbiotic plasmid, pSym, nodulation, transcription
INTRODUCTION
In Rhixobizdm species, more than fifty different genes that
participate in nodulation and nitrogen fixation have been
identified (Fischer, 1994; Schultze e t al., 1994; Van Rhijn
& Vanderleyden, 1995). Current research is mostly
devoted to clarifying the role of specific genes in definite
steps of the Rhipbiztm-legume interaction (Carlson e t ul.,
1994; Schultze e t al,, 1994). In several Rbixobinm species,
most of these genes are clustered in a sector of a single
large plasmid, the so-called symbiotic plasmid or pSym.
Complete physical maps of the pSyms of some Rhixobizdm
species have been published (Girard e t a/., 1991 ; Perret e t
al., 1991); genes involved in the symbiotic process cover
Abbreviation: 32P-sscDNA, 32P-labelled single-stranded complementary
DNA.
a relatively small zone of these megaplasmids. It is
commonly assumed that some of the remaining genetic
information is required for stable maintenance of the
plasmid, or that it participates in other, as yet undescribed,
processes.
The identification of many symbiotic genes has been
achieved mainly through loss of function approaches,
such as random transposon mutagenesis and/or gene
fusion techniques. Cloning and sequencing of interesting
regions and their surroundings has been useful, due to the
observed clustering of genes involved in the symbiotic
process (Sanjuan e t ul., 1993). However, application of
these approaches to the identification of regions responding to different environmental stimuli usually requires
new mutant hunts or extensive sequencing efforts. Moreover, these strategies usually do not detect genes with a
leaky phenotype when mutated. This situation may arise
2847
0002-09230 1996 SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51
L. G I R A R D a n d OTHERS
either when active reiterations (a common situation in
Rhiq+irn species) or alternative pathways occur. These
strategies may also fail to detect genes with an ancillary,
albeit important, role in the symbiotic interaction. These
limitations probably result in an underestimation of the
actual number of genes participating in bacteria-plant
interactions.
follows. The strain CFNX248 was constructed by a triparental
mating using R. etli CFNX89 as the recipient and HB101/
pNC206 (Table 1) as cointegrative plasmid for HB101/pGC479 ;
cointegration between these plasmids occurred through recombination among ampicillin resistance genes shared by these
plasmids. The derivatives were selected as Nal' Cm' Km'
transconjugants. They were analysed by Southern blot hybridization and plasmid profiles (Eckhardt, 1978).
A global approach for the detection of transcriptional
activity under specific conditions offers an alternative to
circumvent the above problems. This strategy relies on
the quantitative detection of regions undergoing active
transcription under diverse experimental conditions. Application of this strategy to sectors of the genome where
a comprehensive map is available allows a positional
evaluation of gene activity in response to the stimulus to
be tested (Chuang e t al., 1993). A similar approach was
recently employed on the pSym of Rhizobimz sp. NGR234
(Fellay etal., 1995) to identify regions whose expression is
enhanced by plant root exudates or specific flavonoids.
Growth conditions. Cultures were grown to mid-exponential
phase in PY medium. Cells were collected by centrifugation (at
6000g in a Sorvall SS34 rotor), washed with sterile minimal
medium and concentrated 100-fold. Fresh Y minimal medium
containing the desired carbon and nitrogen sources was
inoculated with these suspensions at an initial OD,,, of 0.05.
For growth under microaerobic conditions, sodiu-m succinate
and ammonium chloride were used as carbon and nitrogen
sources, respectively. Twenty milliliters of these cultures were
incubated either in 150 ml bottles closed with a cotton stopper
(aerobic conditions), or in 150 ml bottles closed with an airtight
stopper and previously flushed with several volumes of an
oxygen/argon (1 :99, v/v) mixture (microaerobic conditions).
Both types of cultures were grown with shaking (200 r.p.m.) for
8 h at 30 OC. Cells were then collected by centrifugation (6000 g)
at 4 "C and pellets were stored at - 70 O C until RNA isolation.
The use of this technique, coupled with the existence of a
complete map of the pSym of Rhixobiz4m etli (Girard e t al.,
1991), allows us to evaluate the extent of transcriptional
activity under different conditions. In this work, we
report quantitative maps of gene expression in the pSym
in response to growth on several carbon and nitrogen
sources, flavonoid induction and microaerobic conditions, and during symbiosis.
METHODS
Bacterial strains, plasmids and media. The bacterial strains
and plasmids used are listed in Table 1. RhiTobizun strains were
grown at 30 "C in PY rich medium (Noel e t al., 1984) or in Y
minimal medium with different carbon and nitrogen sources,
each at a concentration of 10 mM (Bravo & Mora, 1988).
Carbon sources were sodium succinate, glucose or glycerol;
nitrogen sources were ammonium chloride, potassium nitrate,
sodium glutamate or glutamine. Escherichia coli strains were
grown at 37 "C in LB medium (Miller, 1972). Antibiotics were
added at the following concentrations : chloramphenicol (Cm),
25 pg ml-' (E. colz) or 10 pg ml-' (R. etlz]; kanamycin (Km),
30 pg ml-' (R. etlz]; nalidixic acid (Nal), 20 pg ml-' (R. etlz);
spectinomycin (Sp), 50 pg ml-' (E. coli or R. etlz); tetracycline,
10 pg ml-' (E. colz) or 5 pg ml-' (R. etlz].
Genetic manipulations. Plasmid pGC479 is a pSUP202 derivative (Simon e t al., 1983) that carries a 2170 bp Hind111
fragment containing the nifAgene of R. etli CFN42 (GenBank
accession no. U31630). As a first step in the generation of a
mutation in the n i f A gene from R. etli, a 394 bp fragment from
nifA (position 1165-1559) was excised from pGC479 by
digestion with EcoRI. A 2.0 kb ClSp/Sm (Prentki & Krisch,
1984) cassette was then ligated in its place, thus generating
pGC480. Double recombinants carrying the nzfA,: :QSp/Sm
allele in R. etli were obtained by standard genetic procedures
(Martinez e t al., 1990). This replacement was confirmed by
Southern hybridization (Ausubel e t al., 1987) using the nifA
gene of R. etli as a probe, and also by plant tests in which this
strain was completely ineffective for nitrogen fixation. The
resulting nifA mutant strain was designated CFNX247.
Construction of a strain lacking the pSym but harbouring the
wild-type nifA gene on a replicating plasmid was done as
Genistein induction. Cultures grown for 1 d on PY agar plates
were used to inoculate a minimal medium with the following
composition: sucrose, 5 mM; KNO,, 1 mM; K,HPO,, 1 mM;
KH,PO,, 7 mM ;MgSO,, 1 mM ;NaCl, 2 mM ;CaCl,, 0 1 mM ;
biotin, 0.4 mM ; ferric citrate, 0.2 mM ; H,BO,, 46 pM ;MnSO,,
9 pM; ZnSO,, 0.076 pM; CuSO,, 0.3 pM; Na,MoO,, 0.4 pM.
Cultures were inoculated at an initial OD,,, of 0.1 and grown
with shaking (200 r.p.m.) for 4 h at 30 "C. Genistein was then
added at a final concentration of 1-2pM, with further incubation
for 1, 4, 8 or 24 h. Control cultures were treated similarly, but
without genistein addition. At the appropriate times, samples
were withdrawn and cells were collected by centrifugation
(6000g) at 4 OC and pellets were stored at -70 "C until RNA
isolation. T o verify the production of nodulation factors,
parallel 1 ml cultures were labelled 3 h after genistein addition
with 0.2 pCi (7.4 kBq) ~-[l-'~C]glucosamine,and grown for
12 h. Culture supernatants were analysed by thin-layer chromatography as described previously ( Mendoza e t al., 1995).
Plant growth. Surface-sterilized Phaseolus vzllgaris cv. Negro
Jamapa seeds (kindly provided by PRONASE) were germinated
under sterile conditions in trays containing vermiculite. After
3 d at 30 OC, the seedlings were transferred to 1 1 plastic pots
filled with sterile vermiculite and inoculated with 1 ml of an
overnight culture (in PY medium) of R. etli CFN42. Plants were
grown in a greenhouse and watered periodically with a N-free
nutrient solution (Vincent, 1970) as previously described
(Romero e t al., 1988). After 25 d, nodules were picked out from
the roots, washed briefly with diethyl-pyrocarbonate-treated
water and surface-sterilized. They were then immediately frozen
in a dry-ice-ethanol bath and stored at - 20 "C until further use.
Bacteroid purification. Bacteroids were purified from root
nodules (3 g, fresh wt), obtained as described above, by
centrifugation through self-generated Percoll gradients (Reibach e t al., 1981). Purified bacteroids were utilized immediately
for RNA purification.
RNA purification. Total RNA extraction from purified bacteroids and from cells grown under aerobic or microaerobic
conditions was done by the acid phenol procedure, employing a
commercial kit (RNaid kit, BIO101). RNA was quantified by
absorbance measurements and electrophoresed under de-
2848
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51
pSym transcriptional patterns in Rhixobim etli
Table 1. Bacterial strains and plasmids used in this study
Strain or plasmid
Strain
Rhixobizrm etli
CFN42
CFNX89
CFNX247
CFNX248
Eschericbia coli
HBlOl
Plasmids
pSUP202
pNC206
pRK2013
pGC479
pGC480
cGD collection
Relevant properties
Source or reference
Wild-type strain
CFN42 derivative lacking the pSym
CFN42 derivative carrying the
nifAA : :nSp/Sm allele
CFNX89 derivative carrying a cointegrate
plasmid pNC2061pGC479 (n;fA+)
Quinto e t al. (1985)
Brom e t al. (1992)
This work
Host for recombinant plasmids
Boyer & RoullandDussoix (1969)
Suicide vector, Ap' Cm' Tc'
A deleted RP4 derivative, Ap' Km', selfreplicating in Rhixobizrm spp.
Conjugation helper
Simon et al. (1983)
A. Piihler, Universitat
Bielefeld, Germany.
Figurski & Helinski
A pSUP202 derivative carrying the wildtype nafA gene
A pGC479 derivative carrying the
nzfAA : : nSp/Sm allele
Ordered cosmid collection from the pSym
of R. e t l i CFN42
This work
naturing conditions in agarose-formaldehyde gels (Ausubel e t
al., 1987).
Synthesis of total 32P-labelledsingle-stranded complementary DNA (32P-sscDNA). This was done by the random
hexamer priming method (Sambrook e t al., 1989). The reaction
mixture contained 1 pg total RNA; 0.3 pg of a random
hexanucleotide (Amersham); 0.5 mM dGTP, dATP and dTTP;
5 pM dCTP; 80 pCi [a-32P]dCTP [6000 Ci mmol-' (222 TBq
mmol-l)] ; 20 units placental ribonuclease inhibitor (Gibco
BRL); 5 mM D T T ; 400 units M-MLV Reverse Transcriptase
(Gibco BRL) in the presence of first strand buffer (Gibco BRL) ;
in a total volume of 30 pl. The reaction mixture was incubated
for 90 min at 37 "C. T o hydrolyse the RNA templates, EDTA,
SDS and NaOH were added at final concentrations of 50 mM,
0.4% and 0.36 M, respectively, and the mixture was incubated
for 30 min at 68 "C. For neutralization, Tris buffer (pH 7.4) and
HC1 were then added at final concentrations of 0.33 M and
0.2 M, respectively. Samples were precipitated with ethanol in
the presence of tRNA as carrier, resuspended in 0.1 M NaOH
and used as hybridization probes (A. Garay, personal communication).
Hybridizationconditions and data analysis. Equal amounts of
the cosmids that cover the whole pSym of R. etli CFN42
(quantified spectrophotometrically) were digested with BamHI,
electrophoresed in 1 % (w/v) agarose gels and transferred to
Nylon membranes (Hybond-N , Amersham) (Ausubel e t al.,
1987). DNA blots were hybridized in a solution containing 50 9'0
(v/v) formamide as described previously (Girard e t al., 1991).
Hybridization signals were integrated by scanning the autoradiographs with a GS300 scanning densitometer using the GS365W data system (Hoefer Scientific Instruments), taking care
that measurements were within the linear range of the film.
+
This work
(1979)
This work
Girard e t al. (1991)
RESULTS
Experimental design
In order to study the global transcriptional patterns of the
pSym of R. etli CFN42 under different conditions, we
t o o k advantage o f t h e physical m a p o f this plasmid,
previously established i n o u r laboratory. T h i s m a p was
assembled f r o m a n ordered cosmid collection, comprising
85 BamHI fragments, which are numbered consecutively
on the circular m a p f r o m a n arbitrary start in the nzfTrIDK
region a (Girard e t al., 1991). Hybridization with probes
representing the total RNA population f r o m different
conditions against the whole cosmid set allows us t o
determine t h e extent a n d localization o f transcriptional
activity on t h e pSym. To this end, total RNA was isolated
f r o m cultures g r o w n under different conditions and f r o m
bacteroids (see Methods). These R N A s were used as
templates for the synthesis o f 32P-sscDNAs, which were
then employed as probes in hybridization experiments
against t h e whole cosmid collection. In these experiments,
signal intensity in the autoradiographs was taken as a
measure of t h e relative abundance of t h e mRNA corresponding t o a certain region, i.e. its degree of expression. Pairwise comparison between treatments allowed t h e determination o f induction or repression ratios,
which are defined as the quotient between the conditions
t o be compared. To make valid comparisons between the
different data sets, densitometric scannings were normalized according t o the total radioactivity in the probes,
exposure time a n d a m o u n t of cosmid DNA loaded in the
2849
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51
L. G I R A R D a n d O T H E R S
. .
.
........................... ............. ................. ...................................................................................... .... ...................... ....... .... .... ...... .. ... .... .... ..... ....... .............. ....... ... .... ......... ....... ...... ... .... ... .... ... ...... ........ .............. ...
I.
I..
I
.
Fig. I . Detection of transcriptional patterns of the R. etli pSym and their quantification. The patterns were obtained
using as probes 32P-sscDNAsfrom strain CFN42 grown in succinate-ammonium medium under aerobic (a, b) or
microaerobic (c, d) conditions. (a, c) Southern blot hybridizations of a set of cosmid clones from the pSym digested with
BamHI. Lanes: 1, cGD7 (bands 80-2); 2, cGD102 (bands 1-7); 3, cGDlOl (bands 3-12); 4, cGD28 (bands 9-21); 5, cGD15
(bands 20-24); 6, cGD35 (bands 21-26); 7, cGD45 (bands 24-31). (b, d) Results of densitometric integration of
hybridization signals for the whole pSym. Bar numbers indicate the position of specific fragments on the map (Fig. 3a).
gels, as described in Methods. An example of this type of
analysis is shown in Fig. 1.
This experimental design assumes that the signals observed are due to transcription of sequences in the pSym,
which are specifically detected through high-stringency
hybridization against our cosmid collection. Although
this is a reasonable assumption, reiteration of some
potentially transcribable regions occurs in other replicons
in the cell (Girard e t a/., 1991). T o exclude the possibility
that detectable transcription originated from these reiterated regions, RNA was isolated from microaerobic
cultures of strain CFNX248. This strain lacks the pSym
but carries an active nzfA gene on an autonomous, selfreplicating plasmid (see Methods). Microaerobic conditions in the wild-type strain provoked a widespread, n;fAdependent transcription in the pSym (see below). If
detectable transcription originated from reiterated DNA,
hybridization signals should be observed with a probe
made using microaerobically-grown CFNX248. No
hybridization signals were observed with 32P-sscDNA
synthesized from this strain (data not shown). Thus, the
transcriptional activity detected in our experiments
originated from sequences on the pSym, and not from
reiterations located elsewhere in the genome.
Carbon and nitrogen sources modulate
transcriptional activity in the pSym
To evaluate the role of nutritional stimuli on pSym
expression under aerobic conditions, transcript levels
were analysed after growth in a variety of carbon and
nitrogen sources. As shown in Fig. l(a), growth of the
wild-type strain in a C- and N-rich medium (succinate plus
ammonium) was very restrictive in terms of transcriptional activity. Only two regions (bands 30 and 56)
were expressed at a detectable level, while the rest of the
plasmid was not expressed. This profile was taken as the
baseline expression pattern of the pSym. To discern
between effects caused by different N or C sources, media
containing succinate and glutamine, glutamate or KNO,
were used. For comparison of the effects between the
different treatments, expression values obtained in any
given treatment were divided by the corresponding values
2850
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51
pSym transcriptional patterns in Rhixobium etli
obtained under the baseline condition (succinate plus
ammonium), thus deriving induction ratios. Growth on
glutamine (Fig. 2a) led to higher induction than ammonium. In contrast, growth on a moderate (glutamate,
Fig. 2b) or poor (KNO,, Fig. 2c) nitrogen source
provoked a gradual increase in induction ratios. Growth
on KNO, led to the activation of 16 regions, with
induction ratios ranging from 2-5 to 7.3. Growth on
medium containing ammonium and glucose (Fig. 2d) or
glycerol (Fig. 2e) as alternative carbon sources promoted
higher transcriptional activity than succinate, leading to
the activation of 10 regions at induction ratios ranging
from 2.5 to 9.6. Thus, transcriptional activity in the pSym
is modified by the type of nitrogen and carbon sources
employed.
Temporal order of pSym expression in response to
flavonoid induction
Flavonoid compounds are components of root exudates
which are important mediators in the chemical communication between bacteria and leguminous plants.
Flavonoids are responsible for the induction of genes
involved in the nodulation process, but it has been
suggested that they are also involved in other steps of the
symbiotic interaction, such as competitivity (Bhagwat 8c
Keister, 1992; Sadowsky e t al., 1988). T o identify the
extent and location of pSym regions induced by the
isoflavone genistein, cultures were grown in sucroseKNO, minimal medium (see Methods) and expression
values were obtained at different times after genistein
addition. As a control, expression values were obtained
from parallel cultures treated similarly, but without added
genistein. Induction or repression ratios are expressed as
the quotient of the values obtained in the presence vs the
absence of genistein. Four different regions, differing in
temporal order after induction, could be recognized (Fig.
2f-i). Members of region I, spanning bands 1-1 1, showed
an early pattern of induction. Expression was detectable
1 h after genistein addition, achieving maximal levels 24 h
after the onset of the induction period. In contrast,
regions 11-IV (bands 22-26; 43-51 and 80-85, respectively) were clearly induced only after 24 h. Under our
conditions, production of Nod factors was detectable at
15 h after induction (data not shown). All the nodulation
genes so far identified in R.etli are contained in region I
and in part of region IV (see Fig. 3a and Discussion);
however, regions I1 and I11 mark the location of new,
isoflavone-induced sectors that might be involved in
nodulation or other flavonoid-controlled processes.
Transcriptional activity of the pSym is regulated by
oxygen concentration
Oxygen limitation is an important metabolic effector of
expression of many genes involved in nitrogen fixation.
The effect of this environmental variable was explored by
growing the wild-type strain in a medium containing
succinate and ammonium, under both aerobic and microaerobic conditions (see Methods). Extensive expression of
genes borne by the pSym was observed under microaerobic conditions: 67 out of 85 BamHI bands were
expressed (Fig. lb). This is more clearly seen when
induction ratios (microaerobic vs aerobic levels) are
considered. As shown in Fig. 2j, oxygen limitation
provokes widespread induction of expression in the
pSym, with ratios ranging from 2 to 28. Transcription of
most of the n i f a n d j x genes in Rhipbizlm is dependent on
the n i f A gene product, under both microaerobic and
symbiotic conditions (Fischer, 1994). A comparison of the
induction ratios achieved under microaerobic conditions
in the wild-type strain (Fig. 2j) and in the n i f A mutant
strain (Fig. 2k) allowed us to classify the different pSym
regions according to their degree of dependency on the
n z f A gene product. Thirty-seven fragments showed an
absolute dependency for this gene, i.e. they were expressed
in the wild-type strain but not in the n i f A mutant strain.
Twenty-nine fragments showed lower induction ratios in
the n i f A mutant strain, suggesting a partial dependency
for this transcriptional activator. The extent of this
dependency varied widely, since the induction ratios
ranged from 1.2 to values 10-fold larger in the wild-type
as compared to the mutant strain.
These results indicate that (i) oxygen limitation leads to a
dramatic induction in the transcriptional activity of the
pSym, and (ii) that the NifA transcriptional activator
partially controls this response.
General transcriptional activity of the pSym during
symbiosis
The transcription pattern of the pSym during symbiosis
was only studied in the wild-type strain, due to the low
recovery of bacteroids from nodules induced by the n i f A
mutant. To this end, 32P-sscDNA was synthesized from
total bacteroid RNA of the wild-type strain. This 32PsscDNA was then used as a probe. Under symbiotic
conditions, almost all the plasmid was actively transcribed. In fact, the absolute transcriptional levels during
this stage were the highest among the conditions tested.
As shown in Fig. 2(1), induction ratios during symbiosis
vs aerobic conditions (succinate plus ammonium) may be
as high as 46-fold. Two contiguous regions in the pSym,
spanning fragments 80-85 and 1-29, contain most of the
fragments with a high induction during symbiosis (Fig.
21) ; many of the symbiotic genes detected in this organism
are located here (see Discussion).
The induction ratios observed during symbiosis (Fig. 21)
are qualitatively similar to the ones seen during microaerobiosis (Fig. 2j). To further explore the observed
similarity in transcription patterns, data were plotted as
the quotient of the values during symbiosis vs microaerobiosis. As shown in Fig. 2(m), 41 fragments of the
pSym were induced during symbiosis at levels well above
the ones observed under low oxygen conditions. Fourteen
fragments were expressed at the same level in both
conditions ; fifteen fragments that showed expression
under low oxygen tensions were repressed in bacteroids.
2851
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51
L. G I R A R D a n d O T H E R S
15
15
15
s
-
.- 1 0 Y
-
b
D
30
-
56
5
5
1
1
BarnHl fragment
BarnHl fragment
l5
BarnHI fragment
i
:-
K
BarnHl fragment
$
-- 7
E
0
K
W
BamHl fragment
I
101
i
D
-
BarnHi fragment
P l0I
1
5:
-5
:
[-5-
36
W
BarnHl fragment
BarnHl fragment
: :
K
BarnHl fragment
c
I
5OIi
c 30
40
5
25
I
30
;
0
5 20
11
25
10
BarnHl fragment
BarnHl fragment
1
BarnHl fragment
8amHI fragment
.................................. ............................................................
- ..................................................................
......................................................................................................................................
Fig, 2. For legend see facing page.
2852
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51
..,...__.._..,_.,._............
pSym transcriptional patterns in Rhixobim etli
5
I
I 1
20
I I
25
I
I
35
45
I I
I
55
I
I
75
80
I l l
1
65
I
(b)
I
Induction by:
Q
e3
GI utamine
I
I
Glutamate
Nitrate
GI ucose
m!m
G IyceroI
Fig. 3. Transcriptional activity in the R. etli pSym. (a) Linear BarnHl map of the 390 kb pSym of R. etli CFN42, indicating
the location of specific symbiotic genes. The location of these genes was compiled from data given by Girard e t a / . (1991);
Vdzquez et a/. (1991, 1993); and our own unpublished results. Exact locations for specific genes are: nod box, nodA and
nifHDK a (band 1); fixN0 (band 3); fixQP (band 4); nod box, ORF 1-2 and nodB (band 7); nodC (band 8 ) ; nodll (band
10); nodDl (bands 10-1 1); no/€(band 11); nifH c (band 13); fixABCX, nifA and nifB (band 21); nifHDK b (band 25); rnelA
(bands 55-56); nod box, nodD3, nodD2, nod box and nolP (band 85). Numbers indicate the corresponding BarnHl
fragments. (b) Extent of transcriptional activity under each environmental condition tested. In this figure, only bands that
showed an induction ratio over 2.5 were considered.
DISCUSSION
How many regions in a symbiotic plasmid are actively
expressed during symbiosis? How much of this information is regulated by other physiological stimuli? In this
paper we approached these questions through the determination of the global transcription patterns in the
pSym. Our analyses were greatly aided by a sensitive
technique that allows the quantitative detection of the
transcriptional activity in specific genomic regions
(Chuang e t al., 1993). The validity of this approach is
supported by the fact that the transcriptional activity
detected originates from the pSym (see Results).
Two limitations need to be considered to analyse these
experiments. First, it is not generally possible to equate
transcriptional activity of a region with the activity of
single genes due to the possibility that a region may
contain several transcriptional units. A second limitation
is that if transcripts are generated from a multigene family
located exclusively in the same replicon (such as the
nzjHDK reiterations in the pSym), the total transcriptional
activity detected may be due to any or all the members of
the family. Nevertheless, this approach is valuable for the
identification and localization of regions which respond
to defined environmental stimuli.
Fig. 3 shows a correlation between the quantitative maps
of transcriptional activity obtained in this work and the
Fig. 2. Comparison of the levels of induction in the R. etli pSym under different conditions. For each fragment, induction
ratios were calculated as a quotient of absolute transcript levels (Fig. 1) between the conditions to be compared. For
repression, the reciprocal of the induction ratios were plotted. For these comparisons, a value corresponding to the
minimal detection level (AsBo0.01) was assigned t o unexpressed regions. (a-e) Induction by glutamine (a), glutamate (b),
nitrate (c), glucose (d) and glycerol (e) compared with the wild-type strain grown in succinate-ammonium medium under
aerobic conditions. (f-i) Induction by genistein a t 1 h (f), 4 h (g), 8 h (h) and 24 h (i) after addition compared with the
corresponding control without genistein addition. (j, k) Microaerobic induction in the wild-type strain 0) and in the nifA
mutant strain (k) compared with the wild-type strain grown under aerobic conditions. (I, m) Symbiotic induction in the
wild-type strain compared to the wild-type strain grown under aerobic (I) or microaerobic (m) conditions. Bar numbers
indicate the position of specific regions on the map (Fig 3a).
2853
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51
L. G I R A R D a n d O T H E R S
location of specific genes on the pSym. It is noteworthy
that physiological signals, such as the type of the carbon
and nitrogen sources, exert significant control on the
transcriptional activity of the pSym (Fig. 3b). In general,
poorly utilized carbon or nitrogen sources allow a greater
expression than rich ones. The expressed regions appear
to be dispensable, since curing of the pSym does not affect
growth rates under these conditions. Explanta expression
of some symbiotic genes in other Rhixobimz species, such
as nodD3 (Dusha & Kondorosi, 1993), nzfTlDK and
JixABCX (Szeto e t al., 1987), is affected by the nitrogen
source. In fact, it has been proposed that nitrogen status
may control nodulation by R. etli via the ntrC gene
product (Mendoza e t al., 1995). Although the expression
of some fragments under C- or N-limited conditions
might be explained by correspondence with known
symbiotic genes, many of these fragments map in zones of
the pSym lacking identifiable nod or Jix genes. The
functional status of these fragments is still uncertain, but
they represent N- and C-controlled regions that might be
relevant for either saprophytic or symbiotic life-styles.
Four different regions were actively transcribed after
induction with the isoflavone genistein. One of these
(region I, Fig. 2i) contains most of the nodulation genes
identified in R. etli (Fig. 3a). Complementation of a pSymcured strain with a cosmid clone containing region I leads
to nodule formation (Cevallos e t al., 1989); thus, this
region is sufficient for nodulation. Region I shows an
early pattern of expression, detectable 1 h after induction
and achieves maximal levels after 24 h. These data are in
agreement with experiments employing transcriptional
fusions with members of region I, such as nodA and
nodBC, where high-level expression was observed after
24 h (Vizquez e t al., 1791). All the new regions identified
(regions 11-IV, Fig. 2i) show a late pattern of inducibility,
suggesting that these regions might be important in other
flavonoid-influenced processes, such as efficiency of
nodulation (Schultze e t al., 1974), competitivity (Bhagwat
& Keister, 1992; Sadowsky e t al., 1988), chemotaxis
(Aguilar e t al., 1988) o r resistance to plant phytoalexins
(Parniske e t al., 1991).
A similar approach was recently employed with the pSym
of Rhixobim sp. NGR234 (Fellay e t al., 1995). Although
the spatial order of equivalent zones is different from the
pSym of CFN42, four flavonoid-inducible regions were
also observed. Three of these were induced only late after
flavonoid addition. Contrasting with our data, a locus
similar to our region I (containing nodABC and other
host-specificity genes) was expressed only shortly after
flavonoid induction. This region becomes transcriptionally silent late after induction. Thus, both the topological arrangement and the temporal order of expression
of equivalent zones may vary depending on the bacterial
host.
It is commonly thought that the pattern of expression of
nitrogen fixation genes under microaerobic conditions
resembles the pattern observed during endosymbiotic
stages. In fact, it has been recently proposed that oxygen
limitation is a key determinant of the symbiotic pattern of
nitrogen fixation gene expression in alfalfa nodules
(Souptne e t al., 1995). Our results show that oxygen
concentration is one of the main effectors for induction of
transcriptional activity in the pSym. Contrasting with the
low expression of the pSym under most aerobic conditions, microaerobic status leads to a dramatic increase in
the expression level for most of the fragments (Fig. 3b).
Some of the induced fragments correspond to sectors that
contain different nif and Jix genes (for instance, both
fragments 1 and 25 contain n z f i D K operons, fragment 21
contains JixABCX and nifA genes, Fig. 3a). However,
functions encoded in many of the induced fragments are
still unknown.
It is important to note that this microaerobic induction is
dependent, either partially or totally, on the presence of an
active n i f A gene (Fig 2k). This dependency suggests that
many induced fragments might have a role, either essential
or ancillary, during the symbiotic process. Nevertheless, a
detectable fraction of transcriptional activity in this
condition is independent of n i f A . It is possible that the
remaining activity might be dependent on other oxygensensitive regulators, such as JixK (Fischer, 1994).
Symbiotic conditions gave the highest induction in the
pSym. Our results show that most of the plasmid regions
are expressed during symbiosis (Fig. 3b). Two contiguous
regions account for most of the highly expressed fragments in this condition. Although some symbiotic genes,
such as nzJTIDK,JixABCXand nzfA, have been located in
these fragments, the function of the others (Fig. 3a) is
largely unknown.
Considering the level of expression of these regions, it is
plausible to think that they may be particularly relevant
for symbiosis. In support of this, deletions that remove a
sector comprising fragments 1-25 lead to an inability to
nodulate (Romero e t al., 1991). In contrast, deletions that
affect fragments 30-81 are still able to nodulate and to fix
nitrogen (unpublished data). The only other highly
expressed region in the pSym is fragment 55, which
apparently contains the melA gene (unpublished data) ;
this gene is expressed during both microaerobic and
symbiotic stages (Hawkins & Johnston, 1988).
Although our data are consistent with the role of oxygen
as one of the main effectors of symbiotic expression, some
discrepancies still remain. Our data for the symbiotic
induction pattern resemble, but are not identical to, the
microaerobic induction pattern. A comparison of both
conditions (Fig. 2m), revealed that induction ratios were
the same (i.e. close to one) for only fourteen expressed
fragments. The rest of the fragments were either more
induced or even repressed during symbiosis.
Microaerobic cultures similar to the ones we employed are
commonly used to evaluate the level of expression of
known symbiotic genes zn vitro. Our comparisons are
useful to evaluate the agreement between these conditions. Although there is a good qualitative agreement,
there are important differences. It is now well established
that the 0, concentration in the infected nodule tissue is
about 50 nM (Kuzma e t al., 1993). Although the actual
2854
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51
pSym transcriptional patterns in RhZz$~Z~ztn-zetlz
concentration of dissolved oxygen was not determined in
our microaerobic cultures, calculations based on the
published Ostwald coefficients for oxygen suggest that it
was approximately 12 pM. ?'he simplest explanation for
the observed differences bet ween the microaerobic and
symbiotic induction patterns is that they are solely due to
differences in oxygen concentration. Another explanation
is that there are other controls, superimposed on the one
exerted by oxygen concentration, which act to determine
the level of expression during symbiosis. Resolution of
these alternatives must wait for quantitative determinations of expression levels under nanomolar oxygen
concentrations.
Previous attempts to evaluate the global transcriptional
activity of a pSym were restricted to some defined pSym
sectors (David e t a/.,1987) or to isolated highly expressed
regions from genome banks (Scott-Craig e t a/., 1991).
However, a recent report from Fellay e t a/. (1995)
evaluated the transcriptional activity of the whole pSym
of Rhizobitrm sp. NGR234 under flavonoid or root exudate
induction, employing an approach similar to the one
described in this paper. Our data extend the usefulness of
this approach, providing quantitative information on
transcriptional activity on a different pSym under a variety
of environmental conditions.
An additional application of this approach is in the
evaluation of the effect of mutations in global transcriptional regulators, such as nzfA.This type of analysis
allows a rapid evaluation of the number and position of
regions controlled by a complex regulatory circuit, either
in a positive or a negative mode. We believe that the
resolutive power achieved through a quantitative global
approach will make it extremely useful for the analysis of
complex genomes.
Brom, S., Garcia de 10s Santos, A., Stepkowski, T., Flores, M.,
DAvila, G., Romero, D. & Palacios, R. (1992). Different plasmids of
Rhiyobizlm legzlminosarzlm bv. phaseoli are required for optimal
symbiotic performance. J Bacteriol 174, 51 83-51 89.
Carlson, R. W., Price. N. P. J. & Stacey, G.(1994). The biosynthesis
of rhizobial lipo-oligosaccharide nodulation signal molecules. Mol
Plant-Microb Interact 7, 684-695.
Cevallos, M. A., VAzquez, M., Ddvalos, A., Espin, G., Sepdveda, 1.
& Quinto, C. (1989). Characterization of RhiTobinm phaseoli Sym
plasmid regions involved in nodule rnorphogenesis and host-range
specificity. hfol Microbiol3, 879- 889.
Chuang, S., Daniels, D. L. & Blattner, F. R. (1993). Global
regulation of gene expression in Escherichia coii. Bacteriol 175,
2026-2036.
David, M., Domergue, O., Pognonec, P. & Kahn, D. (1987).
Transcription patterns of Rhixobizlm meliloti symbiotic plasmid
pSym : identification of nfA-independent fix genes. J Bacterioll69,
2239-2244.
Dusha, I. & Kondorosi, A. (1993). Genes at different regulatory
levels are required for the ammonia control of nodulation in
RhiTobium meliloti. Mol Gen Genet 240, 435-444.
Eckhardt, T. (1978). A rapid method for the identification of
plasmid deoxyribonucleic acid in bacteria. Plasmid 1, 584-588.
Fellay, R., Perret, X., Viprey, V., Broughton, W. J. & Brenner, S.
(1995). Organization of host-inducible transcripts on the symbiotic
l
657-667.
plasmid of Rhixobiim sp. NGR234. ~ Z l o illicrobioll6,
Figurski, D. H. & Helinski, D. R. (1979). Replication of an origincontaining derivative of plasmid RK2 dependent on a plasmid
i
Sci U S A 76, 1648-1652.
function provided in tram. Proc N ~ tAcad
Fischer, H.-M. (1994). Genetic regulation of nitrogen fixation in
Rhizobia. Microbiol Rev 58, 352-386.
Girard, M. L., Flores, M., Brom, S., Romero, D., Palacios, R. &
DAvila, G. (1991). Structural complexity of the symbiotic plasmid of
Rhipobiidm iegzrmijiosarzlm bv. phaseoli. J Bacteriol 173, 241 1-2419.
Martinez-Salazar for critical reviewing of the manuscript ; t o
Adriana Garay (Instituto de Biotecnologia, UNAM) for help
with c D N A synthesis; and t o Sergio Cam, Laura Cervantes and
Maria de la Paz Salas for skilful technical assistance.
Hawkins, F. K. L. & Johnston, A. W. B. (1988). Transcription of a
Rhixobizlm legzlminosarzlm biovar phaseoli gene needed for melanin
synthesis is activated by n z f A of Rhixobizlm and Klebsiellapnezlmoniae.
Mol Microbiol2, 331-337.
Kuzma, M. M., Hunt, 5. & Layzell, D. B. (1993). Role of oxygen in
the limitation and inhibition of nitrogenase activity and respiration
rate in individual soybean nodules. Plant l'bysiol1O1, 161 169.
Martinez, E., Romero, D. & Palacios, R. (1990). The Khiyobizlm
genome. Crit Kez, Plant Sci 9, 59 93.
REFERENCES
Mendoza, A., Leija, A., Martlnez-Romero, E., HernAndez, G. &
Mora, 1. (1995). The enhancement of ammonium assimilation in
We are indebted t o Susana Brom, Patrice Dion and Jaime
legttminosarzlm biovar phaseoli towards flavonoid inducers of the
symbiotic nodulation genes. J Gen Microbioll34, 2741-2746.
Rhiyobizlm etli prevents nodulation of Phaseolzls vzllgaris. Mol PlantMicrobe Interact 8, 584-592.
Miller, J. H. (1972). Experiments in Molertrlar Genetics. Cold Spring
Harbor, NY : Cold Spring Harbor Laboratory.
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman,
J. G., Smith, J. A. & Struhl, K. (1987). Cttrrent Protocols in hfobcular
B i o l ~ Xew
~ . York : John Wiley.
Tn5 insertions. J Bacteriol158, 148-155.
Aguilar, J. M. M., Ashby, A. M., Richards, A. 1. M., Loake, G. A.,
Watson, M. D. & Shaw, C. H. (1988). Chemotaxis of Rhiyobizlm
Noel, K. D., Sanchez, F., Ferndndez, L., Leernans, J. & Cevallos,
M. A. (1984). Rhixobizlmphaseoli symbiotic mutants with transposon
Bhagwat, A. A. & Keister, D. L. (1992). Identification and cloning
Parniske, M., Ahlborn, B. & Werner, D. (1991). Isoflavonoidinducible resistance to the phytoalexin glyceollin in soybean
rhizobia. J Bacteriol 173, 3432-3439.
Boyer, H. W. & Roulland-Dussoix, D. (1969). A complementation
analysis of restriction and modification of D N A in Escherichia coli.
J Mol Biol41, 459-472.
Perret, X., Broughton, W. J. & Brenner, S. (1991). Canonical
ordered cosmid library of the symbiotic plasmid of Rhixobiztm
species NGR234. Prac Xatl A c a d ScI' L'J',4 88, 1923-1927.
Bravo, A. & Mora, J. (1988). Ammonium assimilation in Rhiyobizlm
pbaseoli by the glutamine synthetase-glutamate synthase pathway.
J Bacterioll70, 980-984.
Quinto, C., de la Vega, H., Flores, M., Leemans, J., Cevallos, M. A.,
of Bra&bixobizlm japoniczlm genes expressed strain selectively in soil
and rhizosphere. Appl Enih-on Microbial 58, 1490-1 495.
Prentki, P. & Krisch, H. M. (1984). In uitro insertional mutagenesis
with a selectable DNA fragment. Gene 29, 303-313.
2855
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51
L. G I R A R D a n d O T H E R S
Pardo, M. A., Azpiroz, R., Girard, M. L., Calva, E. & Palacios, R.
(1985). Nitrogenase reductase : a functional multigene family in
Rhixobium phaseoli. Proc Natl Acad Sci U S A 82, 1170-1 174.
Reibach, P. H., Mask, P. L. & Streeter, 1. G. (1981). A rapid one-step
method for the isolation of bacteroids from root nodules of soybean
plants, utilizing self-generating Percoll gradients. Can J Microbiol
27, 491495.
Romero, D., Singleton, P. W., Segovia, L., Morett, E., Bohlool,
B. B., Palacios, R. & Ddvila, G. (1988). Effect of naturally occurring
nzf reiterations on symbiotic effectiveness in Rhixobiztm phaseoli.
Appl Environ Microbiol54, 848-850.
of Bra4rhixobiumjaponiczim DNA sequences that are transcribed at
high levels in bacteroids. Mol Gen Genet 228, 356-360.
Simon, R., Priefer, U. & PUhler, A. (1983). Vector plasmids for invitro manipulations of Gram-negative bacteria. In Molecular Genetics
of the Bacteria-Plant Interaction, pp. 98-106. Edited by A. Piihler.
Berlin : Springer-Verlag.
Souphe, E., Foussard, M., Boistard, P., Truchet, G. & Batut, J.
(1995). Oxygen as a key developmental regulator of Rhiyobizim
meliloti N,-fixation gene expression within the alfalfa root nodule.
Proc Natl Acad Sci U S A 92, 3759-3763.
Szeto, W. W., Nixon, B. T., Ronson, C. W. &Ausubel, F. M. (1987).
Romero, D., Brom, S., Martinez-Salazar, J., Girard, M. L., Palacios,
R. & Ddvila, G. (1991). Amplification and deletion of a nod-nifregion
Identification and characterization of the Rhixobium meliloti ntrC
gene : R. meliloti has separate regulatory pathways for activation of
nitrogen fixation genes in free-living and symbiotic cells. J Bacteriol
169, 1423-1432.
Sadowsky, M. J., Olson, E. R., Foster, V. E., Kosslak, R. M. &
Verma, D. P. 5. (1988). Two host-inducible genes of Rhixobium
Van Rhijn, P. & Vanderleyden, J. (1995). The Rhixobium-plant
symbiosis. Microbiol Rev 59, 124142.
in the symbiotic plasmid of Rhixobium phaseoli. J Bacteriol 173,
2435-2441.
fredii and characterization of the inducing compound. J Bacteriol
170, 171-178.
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular
Cloning: a Laboratoy Manual, 2nd edn. Cold Spring Harbor, NY:
Cold Spring Harbor Laboratory.
Sanjuan, J., Luka, S. & Stacey, G. (1993). Genetic maps of Rhixobium
and Bra4rhixobium species. In Genetic Maps, pp. 2136-2145. Edited
by S. O’Brien. Cold Spring Harbor, NY: Cold Spring Harbor
Laboratory.
Schultze, M., Kondorosi, E., Ratet, P., BuirC, M. & Kondorosi, A.
(1994). Cell and molecular biology of Rhixobium-plant interactions.
VBzquez, M., Ddvalos, A., de las PeAas, A., Sanchez, F. & Quinto,
C. (1991). Novel organization of the common nodulation genes in
Rhixobium leguminosarum bv. phaseoli strains. J Bacteriol 173,
1250-1258
VBzquez, M., Santana, 0. & Quinto, C. (1993). The Nod1 and
Nod J proteins from Rhixobium and Bra4rhixobium strains are
similar to capsular polysaccharide secretion proteins from gramnegative bacteria. Mol Microbiol8, 369-377.
Vincent, 1. M. (1970). A Manualfor the PracticalStu4 of Root Nodule
Bacteria. IBP Handbook no. 15. Oxford: Blackwell Scientific
Publications.
Int Rev Cytol156, 1-75.
Scott-Craig, J. S., Guerinot, M. L. & Chelm, B. K. (1991). Isolation
Received 29 April 1996; revised 19 June 1996; accepted 24 June 1996.
2856
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 11:57:51